So far information retrieval using computers has been in general carried out. In the information retrieval using computers, each processing is executed according to software (programs). Consequently, in the execution of individual processing it is required to take many steps such as the read and interpretation of programs from memory, the read of data from files and the write of calculation results into memory. Thus, even in the simplest procedure, processing time which is five to six times as long as the memory access time is required and the procedures requiring over several-thousand times the access time are not scarce. As a result, despite numerous improvements that have sofar been made to softwares and hardwares, in the retrieval apparatus using computers, those having a satisfactory high speed, that is, those capable of extracting desired information out of voluminous information groups almost instantly have not yet been realized.
In the past, to compensate for the slow-speed software processing, preconstituting the files in the form that conform to the retrieval mode has been generally executed. The files constituted into such a special form is called a data base. However, the data base has such a demerit that its form limits the retrievable range, thus susceptible to overlooking the information. For example, assume that a data base divided into blocks according to each person's name has been configured pursuant to the retrieval mode of extracting the information concerning a person with the person's name as a key. When a retrieval has been conducted with a specific person's name as a key, the block of the person's name will be detected, and the information recorded in the block will be taken out However, because the block of other person's name cannot be discovered, even if the information related to the person keyed in that block has been recorded, it is not possible to find out the information.
Thus, because the conventional information retrieval apparatus conducts the processing using software, a satisfactory high-speed retrieval was not available sofar. Further, to compensate for the low speed, a database of specific form needs to be configured. However, the retrievable range will be restricted thereby, and there was a fear of overlooking the information.
Therefore, the major purpose of this invention is to achieve a satisfactory retrieval of extremely high speed by means of hardware alone without relying on software.
A secondary purpose of this invention is to make the whole range of files retrievable in a short time through an extremely high retrieval speed, thus obviating the need for the data-base fabrication and also eliminating the passibility of overlooking the information. Another secondary purpose of this invention is to make possible the extraction of desired information from volumirous information groups in a short time with an extremely high retrieval speed, thereby enabling simply the realization of a high-grade information processing system, for example, an expert system, a voice recognition apparatus and an automatic translation machine which hitherto needed complex software, without using high-grade softwares.